AMD Signs Letter of Intent With French Government to Power National AI Strategy and Exascale Supercomputer
The multi-year collaboration, signed in Paris on April 16, deepens AMD's role in Alice Recoque — France's planned first exascale supercomputer — and extends to AI research, education, and startup support across the country.
AMD and the French government signed a Letter of Intent in Paris on April 16 to formalize a multi-year collaboration supporting France's National Strategy for Artificial Intelligence. The agreement was signed at the French Ministry of the Economy, Finance and Industrial, Energy and Digital Sovereignty, with participants including senior French government officials and AMD representatives.
The deal centers on AMD's role in Alice Recoque, France's planned first exascale supercomputer and Europe's second, which was announced in November 2025. Alice Recoque will be powered by next-generation AMD EPYC processors codenamed Venice alongside AMD Instinct MI430X GPUs, and is expected to deliver more than one exaflop of high-performance computing performance.
Installation is planned for 2026, with researchers expected to gain access by the end of 2027. The project, representing a total cost of 544 million euros, is funded through EuroHPC JU and the Jules Verne Consortium.
Beyond the supercomputer, the April 16 agreement expands AMD's commitment across three pillars: infrastructure access, research and education, and ecosystem development. AMD will provide hardware, software, and training through its University Program, AI Developer Program, and AI Academy for researchers, developers, and startups operating within France's AI Factory framework.
A dedicated Center of Excellence is planned to provide ongoing expertise and support for Alice Recoque users.
For AMD, the agreement represents more than a hardware supply contract — the company is establishing a deep foothold in a government-backed AI ecosystem at a moment when European nations are competing to develop sovereign AI capabilities that reduce dependence on US hyperscaler infrastructure.
Read the original reporting at AMD.
